In order to insure data integrity through a network, systems often employ redundant hardware. If, however, the network itself is a single point of failure, the redundant hardware is useless and failure will occur without an opportunity for recovery. To remedy this situation, systems of interconnected local area networks have been proposed. In these networks, redundant hardware attaches physically to two different local area networks. If one network becomes a single point of failure, operation switches to the second network. This system, however, is expensive since it increases the number of network cards needed to operate the system. Each device would need an additional network card to tie into the redundant LAN. Therefore, it is desirable to provide fault protection in a local area network system without duplicating hardware for redundancy.